Dirk Smith

Effects of Model Complexity on the Performance of Automated Vehicle Steering Controllers: Model Development, Validation and Comparison

SUMMARY Recent research on autonomous highway vehicles has begun to focus on lateral control strategies. The initial work has focused on vehicle control during low-g maneuvers at constant vehicle speed, typical of lane merging and normal highway driving. In this paper, and its companion paper, to follow, the lateral control of vehicles during high-g emergency maneuvers is addressed. Models of the vehicle dynamics are developed, showing the accuracy of the different models under low and high-g conditions. Specifically, body roll, tire and drive-train dynamics, tire force saturation, and tire side force lag are shown to be important effects to include in models for emergency maneuvers. Current controllers, designed for low-g maneuvers only, neglect these effects. The follow on paper demonstrates the performance of lateral controllers during high-g lateral emergency maneuvers using these vehicle models.

Static output feedback stabilization with prescribed degree of stability

The problem of finding a static output feedback matrix is restated. The new formulation replaces the solution of a set of inversely coupled Lyapunov inequalities with the simultaneous solution to an algebraic Riccati inequality and a Lyapunov inequality. An algorithm is developed based on the restated problem. Unlike previous algorithms, the algorithm is noniterative in linear matrix inequality (LMI) solutions. The algorithm may be used to prescribe a given degree of stability, while keeping the static output feedback gain small. Use of the algorithm is demonstrated via an example.

A non-iterative LMI-based algorithm for robust static-output-feedback stabilization

A linear-matrix-inequality-based (LMI-based) algorithm is developed to stabilize polytopic systems via static output feedback. The algorithm is termed non-iterative in LMI solutions because, unlike previous LMI-based static-outputfeedback algorithms, a finite series of LMI problems are solved without the necessity to repeat the solution to any one of the LMI problems. The algorithm is demonstrated on example problems and numerical experiments borrowed from the literature.

A static-output-feedback design procedure for robust emergency lateral control of a highway vehicle

A linear matrix inequality (LMI)-based procedure for the design of robust static-output-feedback controllers is demonstrated on the problem of emergency lateral control of a highway vehicle with bounded time-varying uncertainties. A linear time-varying (LTV) tire model is used with a yaw-plane model of a highway vehicle to express the problem of emergency lateral control. The vehicle system is parameterized for variation in speed between (15 m/s and 30 m/s) and independent variation of front and rear effective lateral tire stiffnesses (between 30 kN/rad and 60 kN/rad) to form a polytope of linear systems. A stabilizing static-output-feedback controller is designed and its gains are reduced while guaranteeing robust stability.

Effects of Model Complexity on the Performance of Automated Vehicle Steering Controllers: Controller Development and Evaluation

SUMMARY Due to increased traffic congestion and travel times, research in Advanced Vehicle Control Systems (AVCS) has focused on automated lateral and headway control. Automated vehicles are seen as a way to increase freeway capacity and vehicle speeds while reducing accidents due to human error. Recent research in automated lateral control has focused on vehicle control during low-g maneuvers. To increase safety, automated lateral controllers will need to recognize and react to emergency situations. This paper investigates the effects of vehicle and tire model order on the response of automated vehicles to an emergency step lane change using a controller based on linear vehicle and tire models. From these studies it is concluded that control strategies based solely on linear vehicle and tire models are inadequate for emergency vehicle maneuvers. A strategy is then proposed to automatically control vehicles through emergency maneuvers. Here the response of a nonlinear vehicle model is used with a linear s...

Nonlinear-gain-optimized controller development and evaluation for automated emergency vehicle steering

Earlier research in advanced vehicle control systems (AVCS) has focused on automated lateral and headway control during low-g maneuvers. However, most emergency situations involve high-g maneuvers where vehicle performance becomes nonlinear. Robust controllers need to be developed that can react to these emergency situations. This paper investigates the development of a nonlinear-gain-optimized (NGO) controller for automated lateral control during emergencies. The strategy is to use a linear model to define the state model and a nonlinear model to optimize the feedback gains for high-g emergency maneuvers. The performance of the NGO controller is presented at 15 and 30 m/s for a step lane change and a double lane change. The NGO controllers robustness is investigated with respect to changes in tire parameters and the number of passengers.

Output-feedback stabilisation with prescribed degree of stability

The problem of finding a static-output-feedback matrix is restated. The new formulation replaces the solution of a set of inversely-coupled Lyapunov inequalities with the simultaneous solution to an algebraic Riccati inequality and a Lyapunov inequality. An algorithm is developed based on the restated problem. Unlike previous algorithms, the algorithm is non-iterative in linear matrix inequality (LMI) solutions. The algorithm may be used to prescribe a given degree of stability, while keeping the static-output-feedback gain small. Use of the algorithm is demonstrated via two examples.

Eigenvalue assignment optimization for stable and unstable systems

A method for designing linear quadratic regulators (LQR) which meet performance specifications with minimal control-signal amplitude is developed. A non-derivative search routine is used to find a smaller gain vector which places the closed-loop eigenvalues in a specific sector-type region. The method may be used to specify arbitrary damping and settling-time characteristics for both stable and unstable systems. In this paper, controllers are designed for example systems to demonstrate the use of the method. The method is shown to yield smaller gain vectors than a previous method, and may be expanded to address more complicated design problems.